94 



have been so fortunate as to mark down several Eoman sites, 

 and in one of these, on my own farm, an excavation exposed a 

 conple of cart-loads [Putt-loa,ds, Dorset] of stone tiles, of some 

 of the most perfect of which I send yon a paper cutting of their 

 exact size and form. Before, however, I refer more particularly 

 to these points, I would describe the nature of the material of 

 the stone itself. 



Bradford Abbas is situate on the Inferior Oolite and Fullers 

 Earth, but the first of these rocks is — ^here especially — too 

 uneven in its fracture to become fissile, and the Fullers Earth 

 is not superimposed by the Stonesfield Slate, as both this latter 

 and the Great Oolite are absent in Dorset, whilst the Forest 

 Marble, which is not far ofi", is even rougher than that rock 

 near Cirencester. In this position, then, the builders employed 

 a fissile bed from the basement beds of the Lias of the adjoining 

 county — Somerset; and from this source, probably at Sparkford, 

 a material was obtained, which, though tolerably smooth on 

 the surface, yet from its thickness, (1 inch,) is as heavy, if not 

 more so, than the coarser tiles from the Forest Marble. 



The length from the top to the point of the base of Fig. 1 is 

 16 inches ; the breadth in the widest part, 10 inches ; width at 

 the top, 5 inches. 



Now the weight of a tile of this 

 form and size (Fig. 1) is 11 lb.; 

 but if square slabs are used, as 

 as is done by the modems, it would 

 require to be nearly double that 

 weight to effect the same purpose. 

 See Fig. 2. 



These, by being broader at the 

 shoulder than at the top, and made 

 lozenge-shaped at the bottom, are 

 necessarily much Hghter than 

 though they had been slabs of 

 parallelograms, arranged side by 

 side, as is the modern custom. 



Fig. 1.— Eoman Tiles. 



